A professional-grade text-to-speech skill utilizing the Volcengine Doubao Large Model for high-fidelity audio synthesis and voice cloning.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install doubao-tts
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install doubao-tts using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Doubao TTS?
The Doubao TTS skill is a powerful integration for Openclaw Skills that leverages the Volcengine (Bytedance) voice synthesis large models. It enables developers to transform plain text into natural-sounding audio files using advanced HTTP V3 streaming interfaces.
This skill supports a vast library of preset voices (v1.0 and v2.0) ranging from professional news anchors to emotional character roles. It also features a specialized mode for voice cloning, allowing users to synthesize audio using custom voice IDs (prefixed with S_) for personalized applications.
Doubao TTS Use Cases
- Creating high-quality audio content for podcasts, videos, or audiobooks.
- Implementing realistic AI voice assistants with specific emotional tones.
- Generating personalized messages using custom voice cloning profiles.
- Localized content creation with support for multiple Chinese dialects and English accents.
- Automating text-to-speech workflows for educational or customer service platforms.
How Doubao TTS Works
- The system validates the required credentials (APP ID and Access Token) from the environment or input parameters.
- The user provides text content and selects a specific speaker ID or emotional profile.
- The skill determines the correct resource_id based on the selected voice version (e.g., seed-tts-2.0 for v2.0 voices).
- A streaming HTTP request is sent to the Doubao TTS API, which returns base64 encoded audio data.
- The script automatically decodes and concatenates the audio stream into a final file (MP3, OGG, or PCM).
- The generated file is saved to the specified output path and presented to the user.
Doubao TTS Setup
First, obtain your credentials from the Volcengine Console (Doubao Voice section). Set up your environment as follows:
export DOUBAO_APP_ID="your_app_id"
export DOUBAO_ACCESS_KEY="your_access_token"
# Optional: Set a default speaker
export DOUBAO_SPEAKER="zh_female_xiaohe_uranus_bigtts"
To synthesize audio via CLI:
python3 scripts/tts_synthesize.py \
--text "Hello world" \
--output output.mp3
Doubao TTS Data Schema & Taxonomy
The skill manages parameters and outputs through a structured schema to ensure compatibility with Openclaw Skills:
| Parameter | Type | Description |
|---|---|---|
| text | String | The source text to be converted (Max 5000 chars recommended) |
| speaker | String | Voice ID (e.g., zh_female_xiaohe_uranus_bigtts or custom S_xxx) |
| emotion | String | Optional emotional tone (happy, sad, angry, etc.) |
| format | Enum | Output file format: mp3, ogg_opus, or pcm |
| sample_rate | Integer | Audio quality setting (default 24000) |
Generated files are typically stored in /mnt/user-data/outputs/.
Doubao TTS Advanced Features
- Emotional Synthesis: Control emotional intensity (scale 1-5) for supported v1.0 and v2.0 voices.
- Voice Cloning Support: Seamless integration with the Volcengine Voice Cloning console for custom S-prefix IDs.
- Automatic Resource Mapping: Intelligently switches between seed-tts-1.0, seed-tts-2.0, and seed-icl-1.0 resource IDs based on the selected voice.
- Session Management: Uses requests.Session for optimized connection reuse in multi-request workflows.
- Long Text Handling: Capability to split and concatenate text segments for content exceeding character limits.
SKILL.md
Loading
Related Openclaw Skills
Bazi Fortune Analysis
A professional-grade BaZi destiny analyzer based on the Ziping Zhenquan system for precise Four Pillars calculations and LLM-powered life path interpretation.
Aider
Aider is a terminal-based AI pair programming tool that enables direct editing of local Git repositories through natural language dialogue.
Tweet Share Card
An automated utility that transforms X/Twitter URLs into polished, high-resolution share cards with customizable aesthetic backgrounds.
OIDC Integration
A specialized skill for designing and implementing secure OpenID Connect and OAuth 2.0 authentication flows across modern web stacks.
Pyramid Principle
A strategic thinking skill that applies the MECE principle to break down complex problems and communicate solutions effectively.
Replit Agent
A powerful AI coding assistant specialized in the Replit ecosystem for end-to-end cloud application development.
Featured*
